Key result
Automated U-Net whole heart segmentation in pediatric cardiac CT achieves a ~0.95 Dice similarity coefficient.
Why the study?
This study investigated the usefulness of deep learning methods for segmenting the whole heart region and the cardiac cavity region in pediatric cardiac CT images using U-Net.
Does automated heart segmentation using U-Net provide accurate segmentation in pediatric cardiac CT images?
Population
Pediatric cardiac CT images across four age categories
Comparison
Chambers trained in a lump vs trained separately
Design
Leave-one-subject-out cross-validation study
Authors
Loading...
May facilitate efficient pediatric CT analysis; leaves open prospective clinical validation before adoption.
Does automated heart segmentation using U-Net provide accurate segmentation in pediatric cardiac CT images?
Deep learning using U-Net provides highly accurate automated whole heart segmentation in pediatric cardiac CT across various age groups.
Yoshida et al. (2021) studied Pediatric cardiac CT. U-Net deep learning method was evaluated on Dice similarity coefficient (DSC) for whole heart segmentation. Automated whole heart segmentation using U-Net in pediatric cardiac CT achieved a mean Dice similarity coefficient over 0.95 across all age categories.
Synapse has enriched 4 closely related papers on similar clinical questions. Consider them for comparative context: